oweals/u-boot.git
16 years ago[new uImage] More verbose kernel image uncompress error message
Marian Balakowicz [Wed, 12 Mar 2008 09:33:01 +0000 (10:33 +0100)]
[new uImage] More verbose kernel image uncompress error message

Signed-off-by: Marian Balakowicz <m8@semihalf.com>
16 years ago[new uImage] Use show_boot_progress() for new uImage format
Marian Balakowicz [Wed, 12 Mar 2008 09:33:01 +0000 (10:33 +0100)]
[new uImage] Use show_boot_progress() for new uImage format

This patch allocates a set of show_boot_progress() IDs for new uImage format
and adds show_boot_progress() calls in new uImage format handling code.

Signed-off-by: Marian Balakowicz <m8@semihalf.com>
16 years ago[new uImage] Add new uImage fromat support to fpga command
Marian Balakowicz [Wed, 12 Mar 2008 09:33:01 +0000 (10:33 +0100)]
[new uImage] Add new uImage fromat support to fpga command

Signed-off-by: Marian Balakowicz <m8@semihalf.com>
16 years ago[new uImage] Add new uImage format handling to other bootm related commands
Marian Balakowicz [Wed, 12 Mar 2008 09:33:01 +0000 (10:33 +0100)]
[new uImage] Add new uImage format handling to other bootm related commands

Updated commands:

docboot  - cmd_doc.c
fdcboot  - cmd_fdc.c
diskboot - cmd_ide.c
nboot    - cmd_nand.c
scsiboot - cmd_scsi.c
usbboot  - cmd_usb.c

Signed-off-by: Marian Balakowicz <m8@semihalf.com>
16 years ago[new uImage] Add new uImage format support to imgextract command
Marian Balakowicz [Wed, 12 Mar 2008 09:33:00 +0000 (10:33 +0100)]
[new uImage] Add new uImage format support to imgextract command

Signed-off-by: Marian Balakowicz <m8@semihalf.com>
16 years ago[new uImage] Add new uImage format support to autoscript routine
Marian Balakowicz [Wed, 12 Mar 2008 09:33:00 +0000 (10:33 +0100)]
[new uImage] Add new uImage format support to autoscript routine

autoscript() routine is updated to accept second argument, which
is only used for FIT images and provides a FIT subimage unit name.

autoscript() routine callers must now pass two arguments. For
non-interactive use (like in cmd_load.c, cmd_net.c), new environment
variable 'autoscript_uname' is introduced and used as a FIT
subimage unit name source.

autoscript command accepts extended syntax of the addr argument:
addr:<subimg_uname>

Signed-off-by: Marian Balakowicz <m8@semihalf.com>
16 years ago[new uImage] Add new uImage format support to arch specific do_bootm_linux() routines
Marian Balakowicz [Wed, 12 Mar 2008 09:33:00 +0000 (10:33 +0100)]
[new uImage] Add new uImage format support to arch specific do_bootm_linux() routines

This patch updates architecture specific implementations of
do_bootm_linux() adding new uImage format handling for
operations like get kernel entry point address, get kernel
image data start address.

Signed-off-by: Marian Balakowicz <m8@semihalf.com>
16 years ago[new uImage] Add node offsets for FIT images listed in struct bootm_headers
Marian Balakowicz [Wed, 12 Mar 2008 09:32:59 +0000 (10:32 +0100)]
[new uImage] Add node offsets for FIT images listed in struct bootm_headers

This patch adds new node offset fields to struct bootm_headers
and updates bootm_headers processing code to make use of them.
Saved node offsets allow to avoid repeating fit_image_get_node() calls.

Signed-off-by: Marian Balakowicz <m8@semihalf.com>
16 years ago[new uImage] ppc: Add new uImage format support to FDT handling routines
Marian Balakowicz [Wed, 12 Mar 2008 09:32:53 +0000 (10:32 +0100)]
[new uImage] ppc: Add new uImage format support to FDT handling routines

Support for new (FIT) format uImages is added to powerpc specific
boot_get_fdt() routine which now recognizes, sanity checks FIT image
and is able to access data sections of the requested component image.

Signed-off-by: Marian Balakowicz <m8@semihalf.com>
16 years ago[new uImage] Re-enable interrupts for non automatic booting
Marian Balakowicz [Wed, 12 Mar 2008 09:14:57 +0000 (10:14 +0100)]
[new uImage] Re-enable interrupts for non automatic booting

Re-enable interrupts if we return from do_bootm_<os> and 'autostart'
environment variable is not set to 'yes'.

Signed-off-by: Marian Balakowicz <m8@semihalf.com>
16 years ago[new uImage] Remove unnecessary arguments passed to ramdisk routines
Marian Balakowicz [Wed, 12 Mar 2008 09:14:38 +0000 (10:14 +0100)]
[new uImage] Remove unnecessary arguments passed to ramdisk routines

boot_get_ramdisk() and image_get_ramdisk() do not need all
cmdtp, flag, argc and argv arguments. Simplify routines definition.

Signed-off-by: Marian Balakowicz <m8@semihalf.com>
16 years ago[new uImage] Add new uImage format support for ramdisk handling
Marian Balakowicz [Wed, 12 Mar 2008 09:12:37 +0000 (10:12 +0100)]
[new uImage] Add new uImage format support for ramdisk handling

This patch updates boot_get_ramdisk() routine adding format
verification and handling for new (FIT) uImages.

Signed-off-by: Marian Balakowicz <m8@semihalf.com>
16 years ago[new uImage] Add new uImage format support for kernel booting
Marian Balakowicz [Wed, 12 Mar 2008 09:01:05 +0000 (10:01 +0100)]
[new uImage] Add new uImage format support for kernel booting

New format uImages are recognized by the bootm command,
validity of specified kernel component image is checked and
its data section located and used for further processing
(uncompress, load, etc.)

Signed-off-by: Marian Balakowicz <m8@semihalf.com>
16 years ago[new uImage] Add new uImage format support for imls and iminfo commands
Marian Balakowicz [Tue, 11 Mar 2008 11:35:20 +0000 (12:35 +0100)]
[new uImage] Add new uImage format support for imls and iminfo commands

imls and iminfo can now recognize nad print out contents of the new (FIT)
format uImages.

Signed-off-by: Marian Balakowicz <m8@semihalf.com>
16 years ago[new uImage] Add support for new uImage format to mkimage tool
Bartlomiej Sieka [Tue, 11 Mar 2008 11:34:47 +0000 (12:34 +0100)]
[new uImage] Add support for new uImage format to mkimage tool

Support for the new uImage format (FIT) is added to mkimage tool.
Commandline syntax is appropriately extended:

mkimage [-D dtc_options] -f fit-image.its fit-image

mkimage (together with dtc) takes fit-image.its and referenced therein
binaries (like vmlinux.bin.gz) as inputs, and produces fit-image file -- the
final image that can be transferred to the target (e.g., via tftp) and then
booted using the bootm command in U-Boot.

Signed-off-by: Bartlomiej Sieka <tur@semihalf.com>
16 years ago[new uImage] Make node unit names const in struct bootm_headers
Marian Balakowicz [Mon, 10 Mar 2008 16:53:49 +0000 (17:53 +0100)]
[new uImage] Make node unit names const in struct bootm_headers

Signed-off-by: Marian Balakowicz <m8@semihalf.com>
16 years ago[new uImage] New uImage low-level API
Marian Balakowicz [Fri, 29 Feb 2008 20:24:06 +0000 (21:24 +0100)]
[new uImage] New uImage low-level API

Add FDT-based functions for handling new format component images,
configurations, node operations, property get/set, etc.

fit_        - routines handling global new format uImage operations
              like get/set top level property, process all nodes, etc.
fit_image_  - routines handling component images subnodes
fit_conf_   - routines handling configurations node

Signed-off-by: Bartlomiej Sieka <tur@semihalf.com>
Signed-off-by: Marian Balakowicz <m8@semihalf.com>
16 years ago[new uImage] Fix FDT blob totalsize calculation in boot_relocate_fdt()
Marian Balakowicz [Fri, 29 Feb 2008 21:22:46 +0000 (22:22 +0100)]
[new uImage] Fix FDT blob totalsize calculation in boot_relocate_fdt()

Do not use global fdt blob pointer, calculate blob size from routine
argument blob pointer.

Signed-off-by: Marian Balakowicz <m8@semihalf.com>
16 years agolibfdt: Add and use a node iteration helper function.
David Gibson [Mon, 11 Feb 2008 13:58:31 +0000 (00:58 +1100)]
libfdt: Add and use a node iteration helper function.

This patch adds an fdt_next_node() function which can be used to
iterate through nodes of the tree while keeping track of depth.  This
function is used to simplify the iteration code in a lot of other
functions, and is also exported for use by library users.

Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
16 years ago[new uImage] Add libfdt support to mkimage
Bartlomiej Sieka [Fri, 29 Feb 2008 15:00:24 +0000 (16:00 +0100)]
[new uImage] Add libfdt support to mkimage

Signed-off-by: Bartlomiej Sieka <tur@semihalf.com>
16 years ago[new uImage] Add sha1.o object to mkimage binary build
Bartlomiej Sieka [Fri, 29 Feb 2008 15:00:23 +0000 (16:00 +0100)]
[new uImage] Add sha1.o object to mkimage binary build

Signed-off-by: Bartlomiej Sieka <tur@semihalf.com>
16 years ago[new uImage] Fix component handling for legacy multi component images
Marian Balakowicz [Fri, 29 Feb 2008 15:00:06 +0000 (16:00 +0100)]
[new uImage] Fix component handling for legacy multi component images

Use uint32_t when accessing size table in image_multi_count() and
image_multi_getimg() for multi component images.

Add missing uimage_to_cpu() endianness conversion.

Signed-off-by: Marian Balakowicz <m8@semihalf.com>
16 years ago[new uImage] Share common uImage code between mkimage and U-boot
Marian Balakowicz [Fri, 29 Feb 2008 14:59:59 +0000 (15:59 +0100)]
[new uImage] Share common uImage code between mkimage and U-boot

This patch adds the following common routines:

1) Dedicated mkimage print_header() is replaced with common
image_print_contents()
image_print_contents_noindent()

2) Common os/arch/type/comp fields name <--> id translation routines
genimg_get_os_name()
genimg_get_arch_name()
genimg_get_type_name()
genimg_get_comp_name()
genimg_get_os_id()
genimg_get_arch_id()
genimg_get_type_id()
genimg_get_comp_id()

Signed-off-by: Marian Balakowicz <m8@semihalf.com>
16 years ago[new uImage] Update naming convention for bootm/uImage related code
Marian Balakowicz [Fri, 29 Feb 2008 13:58:34 +0000 (14:58 +0100)]
[new uImage] Update naming convention for bootm/uImage related code

This patch introduces the following prefix convention for the
image format handling and bootm related code:

genimg_ - dual format shared code
image_ - legacy uImage format specific code
fit_ - new uImage format specific code
boot_ - booting process related code

Related routines are renamed and a few pieces of code are moved around and
re-grouped.

Signed-off-by: Marian Balakowicz <m8@semihalf.com>
16 years agoMerge branch 'master' of git://www.denx.de/git/u-boot into new-image
Marian Balakowicz [Fri, 29 Feb 2008 12:56:44 +0000 (13:56 +0100)]
Merge branch 'master' of git://denx.de/git/u-boot into new-image

16 years ago[new uImage] Respect autostart setting in linux bootm
Kumar Gala [Thu, 28 Feb 2008 03:51:51 +0000 (21:51 -0600)]
[new uImage] Respect autostart setting in linux bootm

Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Acked-by: Marian Balakowicz <m8@semihalf.com>
16 years ago[new uImage] Provide ability to restrict region used for boot images
Kumar Gala [Thu, 28 Feb 2008 03:51:50 +0000 (21:51 -0600)]
[new uImage] Provide ability to restrict region used for boot images

Allow the user to set 'bootm_low' and 'bootm_size' env vars as a way
to restrict what memory range is used for bootm.

Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Acked-by: Marian Balakowicz <m8@semihalf.com>
16 years ago[new uImage] Use lmb for bootm allocations
Kumar Gala [Thu, 28 Feb 2008 03:51:49 +0000 (21:51 -0600)]
[new uImage] Use lmb for bootm allocations

Convert generic ramdisk_high(), get_boot_cmdline(), get_boot_kbd()
functions over to using lmb for allocation of the ramdisk, command line
and kernel bd info.

Convert PPC specific fdt_relocate() to use lmb for allocation of the device
tree.

Provided a weak function that board code can call to do additional
lmb reserves if needed.

Also introduce the concept of bootmap_base to specify the offset in
physical memory that the bootmap is located at.  This is used for
allocations of the cmdline, kernel bd, and device tree as they should
be contained within bootmap_base and bootmap_base + CFG_BOOTMAPSZ.

Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
16 years ago[new uImage] Add autostart flag to bootm_headers structure
Kumar Gala [Thu, 28 Feb 2008 03:51:48 +0000 (21:51 -0600)]
[new uImage] Add autostart flag to bootm_headers structure

The autostart env variable was dropped as part of the initial new uImage
cleanup.  Add it back here so the arch specific code can decide if it
wants to really boot or not.

Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Acked-by: Marian Balakowicz <m8@semihalf.com>
16 years ago[new uImage] Introduce lmb from linux kernel for memory mgmt of boot images
Kumar Gala [Thu, 28 Feb 2008 03:51:47 +0000 (21:51 -0600)]
[new uImage] Introduce lmb from linux kernel for memory mgmt of boot images

Introduce the LMB lib used on PPC in the kernel as a clean way to manage
the memory spaces used by various boot images and structures.  This code
will allow us to simplify the code in bootm and its support functions.

Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
16 years ago[new uImage] ppc: Allow boards to specify effective amount of memory
Kumar Gala [Wed, 20 Feb 2008 04:03:47 +0000 (22:03 -0600)]
[new uImage] ppc: Allow boards to specify effective amount of memory

For historical reasons we limited the stack to 256M because some boards
could only map that much via BATS.  However newer boards are capable of
mapping more memory (for example 85xx is capable of doing up to 2G).

Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Acked-by: Marian Balakowicz <m8@semihalf.com>
16 years ago[new uImage] rework error handling so common functions don't reset
Kumar Gala [Thu, 28 Feb 2008 03:51:46 +0000 (21:51 -0600)]
[new uImage] rework error handling so common functions don't reset

Changed image_get_ramdisk() to just return NULL on error and have
get_ramdisk() propogate that error to the caller.  It's left to the
caller to call do_reset() if it wants to.

Also moved calling do_reset() in get_fdt() and fdt_relocate() on ppc
to a common location.  In the future we will change get_fdt() and
fdt_relocate() to return success/failure and not call do_reset() at all.

Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Acked-by: Marian Balakowicz <m8@semihalf.com>
16 years ago[new uImage] ppc: Re-order ramdisk/fdt handling sequence
Kumar Gala [Thu, 28 Feb 2008 03:51:45 +0000 (21:51 -0600)]
[new uImage] ppc: Re-order ramdisk/fdt handling sequence

Doing the fdt before the ramdisk allows us to grow the fdt w/o concern
however it does mean we have to go in and fixup the initrd info since
we don't know where it will be.

Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
16 years ago[new uImage] ppc: Determine if we are booting an OF style
Kumar Gala [Thu, 28 Feb 2008 03:51:44 +0000 (21:51 -0600)]
[new uImage] ppc: Determine if we are booting an OF style

If we are bootin OF style than we can skip setting up some things
that are used for the old boot method.

Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Acked-by: Marian Balakowicz <m8@semihalf.com>
16 years ago[new uImage] Don't pass kdb to ramdisk_high since we may not have one
Kumar Gala [Thu, 28 Feb 2008 03:51:43 +0000 (21:51 -0600)]
[new uImage] Don't pass kdb to ramdisk_high since we may not have one

We don't actually need the kdb param as we are just using it to get
bd->bi_memsize which we can get from gd->bd->bi_memsize.  Also, if we
boot via OF we might not actually fill out a kdb.

Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Acked-by: Marian Balakowicz <m8@semihalf.com>
16 years ago[new uImage] Correct raw FDT blob handlig when CONFIG_FIT is disabled
Marian Balakowicz [Wed, 27 Feb 2008 10:02:26 +0000 (11:02 +0100)]
[new uImage] Correct raw FDT blob handlig when CONFIG_FIT is disabled

Dual format image code must properly handle all three FDT passing methods:
- raw FDT blob passed
- FDT blob embedded in the legacy uImage
- FDT blob embedded in the new uImage

This patch enables proper raw FDT handling when no FIT imaeg support
is compiled in. This is a bit tricky as we must dected FIT format even
when FIT uImage handling is not enabled as both FIT uImages and raw FDT
blobs use tha same low level format (libfdt).

Signed-off-by: Marian Balakowicz <m8@semihalf.com>
16 years ago[new uImage] POWERPC: Add image_get_fdt() routine
Marian Balakowicz [Wed, 27 Feb 2008 10:02:26 +0000 (11:02 +0100)]
[new uImage] POWERPC: Add image_get_fdt() routine

FDT blob may be passed either: (1) raw (2) or embedded in the legacy uImage
(3) or embedded in the new uImage. For the (2) case embedding image must be
verified before we get FDT from it. This patch factors out legacy image
specific verification routine to the separate helper routine.

Signed-off-by: Marian Balakowicz <m8@semihalf.com>
Acked-by: Kumar Gala <galak@kernel.crashing.org>
16 years ago[new uImage] Add image_get_kernel() routine
Marian Balakowicz [Wed, 27 Feb 2008 10:02:07 +0000 (11:02 +0100)]
[new uImage] Add image_get_kernel() routine

Legacy image specific verification is factored out to a separate helper
routine to keep get_kernel() generic and simple.

Signed-off-by: Marian Balakowicz <m8@semihalf.com>
Acked-by: Kumar Gala <galak@kernel.crashing.org>
16 years ago[new uImage] Move image verify flag to bootm_headers structure
Marian Balakowicz [Wed, 27 Feb 2008 10:01:04 +0000 (11:01 +0100)]
[new uImage] Move image verify flag to bootm_headers structure

Do not pass image verification flag directly to related routines.
Simplify argument passing and move it to the bootm_header structure which
contains curently processed image specific data and is already being passed
on the argument list.

Signed-off-by: Marian Balakowicz <m8@semihalf.com>
Acked-by: Kumar Gala <galak@kernel.crashing.org>
16 years ago[Makefile] Sort COBJS in lib_<arch> Makefiles
Marian Balakowicz [Wed, 27 Feb 2008 10:00:47 +0000 (11:00 +0100)]
[Makefile] Sort COBJS in lib_<arch> Makefiles

Signed-off-by: Marian Balakowicz <m8@semihalf.com>
16 years ago[new uImage] Optimize gen_get_image() flow control
Marian Balakowicz [Wed, 27 Feb 2008 10:00:47 +0000 (11:00 +0100)]
[new uImage] Optimize gen_get_image() flow control

When CONFIG_HAS_DATAFLASH is not defined gen_get_image() routine has nothing
to do, update its control flow to better reflect that simple case.

Signed-off-by: Marian Balakowicz <m8@semihalf.com>
Acked-by: Kumar Gala <galak@kernel.crashing.org>
16 years ago[new uImage] POWERPC: Split get_fdt() into get and relocate routines
Marian Balakowicz [Mon, 4 Feb 2008 07:28:17 +0000 (08:28 +0100)]
[new uImage] POWERPC: Split get_fdt() into get and relocate routines

PPC specific FDT blob handling code is divided into two separate routines:

get_fdt() - find and verify a FDT blob (either raw or image embedded)
fdt_relocate() - move FDT blob to within BOOTMAP if needed

Signed-off-by: Marian Balakowicz <m8@semihalf.com>
Acked-by: Kumar Gala <galak@kernel.crashing.org>
16 years ago[new uImage] Add dual format uImage support framework
Marian Balakowicz [Mon, 4 Feb 2008 07:28:09 +0000 (08:28 +0100)]
[new uImage] Add dual format uImage support framework

This patch adds framework for dual format images. Format detection is added
and the bootm controll flow is updated to include cases for new FIT format
uImages.

When the legacy (image_header based) format is detected appropriate
legacy specific handling is invoked. For the new (FIT based) format uImages
dual boot framework has a minial support, that will only print out a
corresponding debug messages. Implementation of the FIT specific handling will
be added in following patches.

Signed-off-by: Marian Balakowicz <m8@semihalf.com>
16 years agoCoding style cleanup. Prepare v1.3.2-rc2 release candidate v1.3.2-rc2
Wolfgang Denk [Sun, 24 Feb 2008 14:21:36 +0000 (15:21 +0100)]
Coding style cleanup. Prepare v1.3.2-rc2 release candidate

Signed-off-by: Wolfgang Denk <wd@denx.de>
16 years agoENV: remove saveenv when CFG_ENV_IS_NOWHERE is selected
Jean-Christophe PLAGNIOL-VILLARD [Sat, 23 Feb 2008 11:15:56 +0000 (12:15 +0100)]
ENV: remove saveenv when CFG_ENV_IS_NOWHERE is selected

Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
16 years agoRemove the __STRICT_ANSI__ check from the __u64/__s64 declaration on 32bit targets.
Shinya Kuribayashi [Sat, 23 Feb 2008 08:24:16 +0000 (17:24 +0900)]
Remove the __STRICT_ANSI__ check from the __u64/__s64 declaration on 32bit targets.

----------------------------------------------------------------
Olaf Hering [Wed, 17 Oct 2007 06:27:13 +0000 (23:27 -0700)]

Remove the __STRICT_ANSI__ check from the __u64/__s64 declaration on
32bit targets.

GCC can be made to warn about usage of long long types with ISO C90
(-ansi), but only with -pedantic.  You can write this in a way that even
then it doesn't cause warnings, namely by:

#ifdef __GNUC__
__extension__ typedef __signed__ long long __s64;
__extension__ typedef unsigned long long __u64;
#endif

The __extension__ keyword in front of this switches off any pedantic
warnings for this expression.

Signed-off-by: Olaf Hering <olh@suse.de>
Cc: <linux-arch@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
----------------------------------------------------------------

Signed-off-by: Shinya Kuribayashi <skuribay@ruby.dti.ne.jp>
16 years agocpu/mcf52x2/config.mk: Make needlessly deffered expansions immediate.
Shinya Kuribayashi [Sat, 23 Feb 2008 08:07:57 +0000 (17:07 +0900)]
cpu/mcf52x2/config.mk: Make needlessly deffered expansions immediate.

This will reduce the build time.

Signed-off-by: Shinya Kuribayashi <skuribay@ruby.dti.ne.jp>
16 years agocpu/ppc4xx/config.mk: Make a needlessly deffered expansion immediate.
Shinya Kuribayashi [Sat, 23 Feb 2008 08:05:00 +0000 (17:05 +0900)]
cpu/ppc4xx/config.mk: Make a needlessly deffered expansion immediate.

This will reduce the build time.

Signed-off-by: Shinya Kuribayashi <skuribay@ruby.dti.ne.jp>
16 years agocpu/mips/cofigl.mk: Make a needlessly deffered expansion immediate.
Shinya Kuribayashi [Sat, 23 Feb 2008 07:58:41 +0000 (16:58 +0900)]
cpu/mips/cofigl.mk: Make a needlessly deffered expansion immediate.

This reduces the build time by ~10%. Here's the gth2_config example.

        BEFORE       AFTER
real    0m31.441s    0m27.833s
user    0m24.766s    0m23.045s
sys     0m10.425s    0m7.468s

Signed-off-by: Shinya Kuribayashi <skuribay@ruby.dti.ne.jp>
16 years agomake define2mk.sed work on FreeBSD
Marcel Moolenaar [Fri, 22 Feb 2008 18:48:07 +0000 (10:48 -0800)]
make define2mk.sed work on FreeBSD

In the thread "[1.3.2-rc1] MPC8548CDS/MPC8555CDS configs fails to link",
the define2mk.sed script was identified as the source of the link
failure on FreeBSD. The problem is that sed(1) does not always support
the '+' operator. It isn't on FreeBSD. The attach patch implements the
equivalent, using the '*' operator instead and should work everywhere.

Signed-off-by: Marcel Moolenaar <marcelm@juniper.net>
16 years agoReplace deprecated "ramdisk" with "ramdisk_size" kernel parameter.
Detlev Zundel [Fri, 22 Feb 2008 16:21:32 +0000 (17:21 +0100)]
Replace deprecated "ramdisk" with "ramdisk_size" kernel parameter.

The Linux commit fac8b209b1084bc85748bd54e13d00c1262b220f ("Remove
final traces of long-deprecated "ramdisk" kernel parm") makes these
changes neccessary.

Signed-off-by: Detlev Zundel <dzu@denx.de>
16 years agoLM75 bug fix for negative temperatures
Larry Johnson [Thu, 21 Feb 2008 18:58:16 +0000 (13:58 -0500)]
LM75 bug fix for negative temperatures

When the LM75 temperature sensor measures a temperature below 0 C, the
current driver does not perform sign extension, so the result returned is
256 C too high.  This patch fixes the problem.

Signed-off-by: Larry Johnson <lrj@acm.org>
16 years agoIDS8247: update MAINTAINER entry.
Heiko Schocher [Thu, 21 Feb 2008 17:33:45 +0000 (18:33 +0100)]
IDS8247: update MAINTAINER entry.

Signed-off-by: Heiko Schocher <hs@denx.de>
16 years agoFix device tree for mgsuvd board.
Heiko Schocher [Thu, 21 Feb 2008 17:31:15 +0000 (18:31 +0100)]
Fix device tree for mgsuvd board.

Rename the "scc" node in "ethernet" for the mgsuvd board.

Signed-off-by: Heiko Schocher <hs@denx.de>
16 years agolwmon5: enable hardware watchdog
Yuri Tikhonov [Thu, 21 Feb 2008 13:23:42 +0000 (14:23 +0100)]
lwmon5: enable hardware watchdog

Some boards (e.g. lwmon5) may use rather small watchdog intervals, so
causing it to reboot the board if U-Boot does a long busy-wait with
udelay(). Thus, for these boards we have to restart WD more
frequently.

This patch splits the busy-wait udelay() into smaller, predefined,
intervals, so that the watchdog timer may be resetted with the
configurable (CONFIG_WD_PERIOD) interval.

Signed-off-by: Yuri Tikhonov <yur@emcraft.com>
16 years agoppc4xx: Support for ATI Radeon 9200 card on sequoia
Anatolij Gustschin [Thu, 21 Feb 2008 11:52:29 +0000 (12:52 +0100)]
ppc4xx: Support for ATI Radeon 9200 card on sequoia

Adds configuration option for ATI Radeon 9200 card
support to sequoia config file. If CONFIG_VIDEO
is enabled, TEXT_BASE should be changed to 0xFFF80000.

Signed-off-by: Anatolij Gustschin <agust@denx.de>
16 years agoMerge branch 'master' of git://www.denx.de/git/u-boot-mpc86xx
Wolfgang Denk [Fri, 22 Feb 2008 12:03:28 +0000 (13:03 +0100)]
Merge branch 'master' of git://denx.de/git/u-boot-mpc86xx

16 years agoMerge branch 'master' of git://www.denx.de/git/u-boot-cfi-flash
Wolfgang Denk [Fri, 22 Feb 2008 12:02:10 +0000 (13:02 +0100)]
Merge branch 'master' of git://denx.de/git/u-boot-cfi-flash

16 years agoMerge branch 'master' of git://www.denx.de/git/u-boot-freebsd
Wolfgang Denk [Fri, 22 Feb 2008 12:01:04 +0000 (13:01 +0100)]
Merge branch 'master' of git://denx.de/git/u-boot-freebsd

16 years agoMerge branch 'master' of git://www.denx.de/git/u-boot-coldfire
Wolfgang Denk [Fri, 22 Feb 2008 11:56:11 +0000 (12:56 +0100)]
Merge branch 'master' of git://denx.de/git/u-boot-coldfire

16 years agoRemove duplicate defines for ARRAY_SIZE
Kumar Gala [Mon, 18 Feb 2008 14:18:07 +0000 (08:18 -0600)]
Remove duplicate defines for ARRAY_SIZE

A few duplicate of the ARRAY_SIZE macro sneaked in since we put
the define in common.h.

Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
16 years agoppc: Allow boards to specify effective amount of memory
Kumar Gala [Mon, 18 Feb 2008 14:09:37 +0000 (08:09 -0600)]
ppc: Allow boards to specify effective amount of memory

For historical reasons we limited the stack to 256M because some boards
could only map that much via BATS.  However newer boards are capable of
mapping more memory (for example 85xx is capable of doing up to 2G).

Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
16 years agoinclude autoconf.mk before any other .mk files
Mike Frysinger [Mon, 18 Feb 2008 10:24:13 +0000 (05:24 -0500)]
include autoconf.mk before any other .mk files

This bumps the autoconf.mk include step above board/cpu/arch/etc... so that
those .mk files can have make if statements based on the current config.

Signed-off-by: Mike Frysinger <vapier@gentoo.org>
16 years agoerror check autoconf.mk generation
Mike Frysinger [Mon, 18 Feb 2008 10:10:07 +0000 (05:10 -0500)]
error check autoconf.mk generation

If any of the steps for generating autoconf.mk fail currently, they go
unnoticed.  To fix, we can simply add 'set -e' to the long list of commands.
This is simpler and more robust than placing '|| exit $$?' after every line.

Signed-off-by: Mike Frysinger <vapier@gentoo.org>
16 years agoFix bug in dependency checking
Stefano Babic [Mon, 18 Feb 2008 07:03:51 +0000 (08:03 +0100)]
Fix bug in dependency checking

By adding VERSION_FILE to the PHONY targets the script
/tools/setlocalversion is always called and version_autogenerated.h
is replaced only if the script find a modified source file.

Signed-off-by: Stefano Babic <sbabic@denx.de>
16 years agoFix GPMC CS2 memory setup at apollon
Kyungmin Park [Mon, 18 Feb 2008 05:35:43 +0000 (14:35 +0900)]
Fix GPMC CS2 memory setup at apollon

It disables the current map first

Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
16 years agouli526x: Fix multiple differ in signedness and parentheses around comparison
Jean-Christophe PLAGNIOL-VILLARD [Sun, 17 Feb 2008 22:52:46 +0000 (23:52 +0100)]
uli526x: Fix multiple differ in signedness and parentheses around comparison

Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
16 years agoMIPS: Fix CFG_NO_FLASH support
Jean-Christophe PLAGNIOL-VILLARD [Sun, 17 Feb 2008 15:58:04 +0000 (16:58 +0100)]
MIPS: Fix CFG_NO_FLASH support

- Fix flash_init call when CFG_NO_FLASH is used
- Remove no more needed flash.c for qemu-mips

Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
16 years agoeasylogo: clean up some more and add -r (rgb) support
Mike Frysinger [Sat, 16 Feb 2008 07:40:18 +0000 (02:40 -0500)]
easylogo: clean up some more and add -r (rgb) support

Michael Hennerich added support for outputting an image in RGB format rather
than forcing YUYV all the time.  This makes obvious sense if the display you
have takes RGB input rather than YUYV.

Rather than hack in support for options, I've converted it to use getopt and
cleaned up the argument parsing in the process.

Signed-off-by: Michael Hennerich <michael.hennerich@analog.com>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
16 years agoMakefile: add target for $(LDSCRIPT)
Mike Frysinger [Sat, 16 Feb 2008 07:12:37 +0000 (02:12 -0500)]
Makefile: add target for $(LDSCRIPT)

If the $(LDSCRIPT) does not exist (normally it's board/$(BOARD)/u-boot.lds),
then change into the board directory and try and create it.  This allows you
to generate the linker script on the fly based upon board defines (like the
Blackfin boards do).

There should be no regressions due to this change as the normal case is to
already have a u-boot.lds file.  If that's the case, then there's nothing to
generate, and so make will always exit.  The fix here is that if the linker
script does not exist, the implicit rules take over and attempt to guess how
to generate the file.

Signed-off-by: Mike Frysinger <vapier@gentoo.org>
16 years ago[new uImage] Fix erroneous use of image_get_magic() in fdc/usb cmds
Marian Balakowicz [Thu, 21 Feb 2008 16:27:49 +0000 (17:27 +0100)]
[new uImage] Fix erroneous use of image_get_magic() in fdc/usb cmds

Signed-off-by: Marian Balakowicz <m8@semihalf.com>
16 years ago[new uImage] Rename and move print_image_hdr() routine
Marian Balakowicz [Thu, 21 Feb 2008 16:27:41 +0000 (17:27 +0100)]
[new uImage] Rename and move print_image_hdr() routine

Signed-off-by: Marian Balakowicz <m8@semihalf.com>
16 years ago[new uImage] Add fit_parse_conf() and fit_parse_subimage() routines
Marian Balakowicz [Thu, 21 Feb 2008 16:20:20 +0000 (17:20 +0100)]
[new uImage] Add fit_parse_conf() and fit_parse_subimage() routines

Introducing routines for parsing new uImage format bootm arguments:
[<addr>]#<conf> - configuration specification
[<addr>]:<subimg> - subimage specification

New format images can contain multiple subimages of the same type. For example
a single new format image file can contain three kernels, two ramdisks and a
couple of FDT blobs. Subimage and configuration specifications are extensions
to bootm (and other image-related commands) arguments' syntax that allow to
specify which particular subimage should be operated on.

Subimage specification is used to denote a particular subimage. Configurations
are a bit more complex -- they are used to define a particualr booting setup,
for example a (kernel, fdt blob) pair, or a (kernel, ramdisk, fdt blob) tuple,
etc.

Signed-off-by: Marian Balakowicz <m8@semihalf.com>
16 years ago[new uImage] Add gen_get_image() routine
Marian Balakowicz [Thu, 21 Feb 2008 16:20:19 +0000 (17:20 +0100)]
[new uImage] Add gen_get_image() routine

This routine assures that image (whether legacy or FIT) is not
in a special dataflash storage.

If image address is a dataflash address image is moved to system RAM.

Signed-off-by: Marian Balakowicz <m8@semihalf.com>
16 years ago[new uImage] Pull in libfdt if CONFIG_FIT is enabled
Marian Balakowicz [Thu, 21 Feb 2008 16:20:18 +0000 (17:20 +0100)]
[new uImage] Pull in libfdt if CONFIG_FIT is enabled

New uImage format (Flattened Image Tree) requires libfdt
functionality, print out error message if CONFIG_OF_LIBFDT
is not defined.

New uImage support is enabled by defining CONFIG_FIT (and CONFIG_OF_LIBFDT).
This commit turns it on by default.

Signed-off-by: Marian Balakowicz <m8@semihalf.com>
16 years agoMerge branch 'master' of /home/git/u-boot
Marian Balakowicz [Thu, 21 Feb 2008 16:18:01 +0000 (17:18 +0100)]
Merge branch 'master' of /home/git/u-boot

16 years agoCFI: Do not use uninitialized cmd_reset
Michael Schwingen [Mon, 18 Feb 2008 22:16:35 +0000 (23:16 +0100)]
CFI: Do not use uninitialized cmd_reset

Do not use uninitialized cmd_reset; issue both AMD and Intel reset
commands instead

From a short test, it looks like AMD-style flash roms treat *any* unknown
command write as a reset, at least when in CFI Query mode, so issuing the
Intel reset command to AMD-style flashs seems safe (from the small sample I
have), plus the 3-cycle magic sequence should kick the state machine into
the right state even without a reset command. Since the AMD-style flashs
require the unlock sequence for real operation, I chose to try the AMD reset
command first, so that Intel flashs do no see an invalid command prior to
the CFI query.

I have tested the patch on AM29LV320-style flashs from Fujitsu and Macronix,
plus Intel StrataFlash.

Signed-off-by: Michael Schwingen <michael@schwingen.org>
Signed-off-by: Stefan Roese <sr@denx.de>
16 years agoAPI: Add (c) and licensing notice to the public API header.
Rafal Jaworowski [Thu, 21 Feb 2008 10:56:44 +0000 (11:56 +0100)]
API: Add (c) and licensing notice to the public API header.

Signed-off-by: Rafal Jaworowski <raj@semihalf.com>
16 years agoFix CPU POST test failure
Yuri Tikhonov [Thu, 21 Feb 2008 10:06:07 +0000 (11:06 +0100)]
Fix CPU POST test failure

The CPU POST test code (run from cpu_post_exec_31()) doesn't follow the
ABI carefully, at least the CR3, CR4, and CR5 fields of CR are clobbered
by it. The gcc-4.2 with its more aggressive optimization exposes this fact.
This patch just saves the CR value before running the test code, so allowing
it to do anything it wants with CR.

Signed-off-by: Dmitry Rakhchev <rda@emcraft.com>
Acked-by: Yuri Tikhonov <yur@emcraft.com>
--

16 years ago8610HPCD: Document the flashbank selection switches.
Jon Loeliger [Wed, 20 Feb 2008 21:26:51 +0000 (15:26 -0600)]
8610HPCD: Document the flashbank selection switches.

Signed-off-by: Jon Loeliger <jdl@freescale.com>
16 years ago86xx: Fix GUR PCI config registers properly.
Jon Loeliger [Wed, 20 Feb 2008 20:22:26 +0000 (14:22 -0600)]
86xx: Fix GUR PCI config registers properly.

Back in commit 975a083a5ef785c414b35f9c5b8ae25b26b41524 where
I tried to "8610HPCD: Fix typos in two PCI setup registers", I
botched it due to not realizing that 8610 and 8641 had different
Global Utility Register defintions, one of which was like 85xx,
and the other wasn't.  Correct this problem by introducing two
symbols, one for each 86xx SoC, but neither of which is named
anything like 85xx.

My bad.  Lovely Wednesday with git bisect.  You know.

Signed-off-by: Jon Loeliger <jdl@freescale.com>
16 years ago8610HPCD: Don't use VIDEO/CFB_CONSOLE by default.
Jon Loeliger [Wed, 20 Feb 2008 18:24:11 +0000 (12:24 -0600)]
8610HPCD: Don't use VIDEO/CFB_CONSOLE by default.

Without an actual supported video card hooked up, enabling
the CONFIG_VIDEO by default just makes it look broken by
routing all console output to the video card.   Don't.

Signed-off-by: Jon Loeliger <jdl@freescale.com>
16 years agoColdFire: Fix missing code flash size for M5485EVB
TsiChungLiew [Wed, 30 Jan 2008 21:08:15 +0000 (15:08 -0600)]
ColdFire: Fix missing code flash size for M5485EVB

Signed-off-by: James Mahan <kmahan@freescale.com>
Signed-off-by: TsiChung Liew <Tsi-Chung.Liew@freescale.com>
16 years agoColdFire: Fix 5282 and 5271 interrupt mask bit
TsiChungLiew [Wed, 30 Jan 2008 21:04:42 +0000 (15:04 -0600)]
ColdFire: Fix 5282 and 5271 interrupt mask bit

Signed-off-by: TsiChung Liew <Tsi-Chung.Liew@freescale.com>
16 years ago8610HPCD: Fix typos in two PCI setup registers.
Jon Loeliger [Tue, 19 Feb 2008 18:31:08 +0000 (12:31 -0600)]
8610HPCD: Fix typos in two PCI setup registers.

The two symbols MPC86xx_PORDEVSR_IO_SEL and MPC86xx_PORBMSR_HA
were erroneously present as 85xx names and values, leftover from
the clone wars.  Fix this by removing the 85xx cruft from the
86xx codebase.

Signed-off-by: Jon Loeliger <jdl@freescale.com>
16 years ago86xx: Convert sbc8641d to use libfdt.
Jon Loeliger [Mon, 18 Feb 2008 20:01:56 +0000 (14:01 -0600)]
86xx: Convert sbc8641d to use libfdt.

This is the proper fix for a missing closing brace in the function
ft_cpu_setup() noticed by joe.hamman <at> embeddedspecialties.com.
The ft_cpu_setup() function in mpc8641hpcn.c should have been
removed earlier as it was under the obsolete CONFIG_OF_FLAT_TREE,
but was missed.  Only, the sbc8641d was nominally still using it.
It all got ripped out, and the funcality that was in ft_board_setup()
was refactored to remove the CPU portions into the new file
cpu/mpc86xx/fdt.c instead.  Make sbc8641d use this now.

Based loosely on an original patch from joe.hamman@embeddedspecialties.com

Signed-off-by: Jon Loeliger <jdl@freescale.com>
16 years agompc86xx: Fix unused variable 'config' and 'immap'
Jean-Christophe PLAGNIOL-VILLARD [Sun, 17 Feb 2008 22:35:31 +0000 (23:35 +0100)]
mpc86xx: Fix unused variable 'config' and 'immap'

and remove useless CONFIG_DDR_INTERLEAVE

Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
16 years agompc86xx: Fix implicit declaration of functions 'init_laws' and 'disable_law'
Jean-Christophe PLAGNIOL-VILLARD [Sun, 17 Feb 2008 22:03:36 +0000 (23:03 +0100)]
mpc86xx: Fix implicit declaration of functions 'init_laws' and 'disable_law'

Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
16 years agos3c24x0: Fix unused variable 'i' in function 'serial_init_dev'
Jean-Christophe PLAGNIOL-VILLARD [Sun, 17 Feb 2008 13:15:31 +0000 (14:15 +0100)]
s3c24x0: Fix unused variable 'i' in function 'serial_init_dev'

Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
16 years agopxa: fix assignment from incompatible pointer type
Jean-Christophe PLAGNIOL-VILLARD [Sun, 17 Feb 2008 13:15:32 +0000 (14:15 +0100)]
pxa: fix assignment from incompatible pointer type

fix mmc_bread function prototype

Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
16 years agoat91cap9adk: fix implicit declaration of function 'eth_init'
Jean-Christophe PLAGNIOL-VILLARD [Sun, 17 Feb 2008 13:15:30 +0000 (14:15 +0100)]
at91cap9adk: fix implicit declaration of function 'eth_init'

Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
16 years agoRemove files added by mistake, update CHANGELOG.
Wolfgang Denk [Sun, 17 Feb 2008 14:43:44 +0000 (15:43 +0100)]
Remove files added by mistake, update CHANGELOG.

Signed-off-by: Wolfgang Denk <wd@denx.de>
16 years agoPPC440EPx: Optionally enable second I2C bus
Mike Nuss [Wed, 6 Feb 2008 16:10:11 +0000 (11:10 -0500)]
PPC440EPx: Optionally enable second I2C bus

The option CONFIG_I2C_MULTI_BUS does not have any effect on Sequoia, the
PPC440EPx reference platform, because IIC1 is never enabled. Add Sequoia board
code to turn on IIC1 if CONFIG_I2C_MULTI_BUS is selected.

Signed-off-by: Mike Nuss <mike@terascala.com>
Cc: Stefan Roese <sr@denx.de>
16 years agoppc4xx: HCU4/5. Cleanup configs
Niklaus Giger [Tue, 5 Feb 2008 09:26:44 +0000 (10:26 +0100)]
ppc4xx: HCU4/5. Cleanup configs

- hcu4.h: Removed define of CONFIG_PPC405GPr
- Corrected phy addresses
- Fix boot variables
- Respect line length of 80 chars

Signed-off-by: Niklaus Giger <niklaus.giger@netstal.com>
16 years agoppc4xx: HCU4/5. Cleanups
Niklaus Giger [Tue, 5 Feb 2008 10:31:28 +0000 (11:31 +0100)]
ppc4xx: HCU4/5. Cleanups

- Fix some coding style violations.
- Use in/out_u16/32 where appropriate.
- Use register names from ppc405.h.
- Fix trace useage for Lauterbach.
- Remove obsolete generation HCU2.
- Renamed fixed_hcu4_sdram to init_ppc405_sdram.

Signed-off-by: Niklaus Giger <niklaus.giger@netstal.com>
16 years agoppc4xx: PPC405GPr fix missing register definitions
Niklaus Giger [Tue, 5 Feb 2008 09:26:41 +0000 (10:26 +0100)]
ppc4xx: PPC405GPr fix missing register definitions

Signed-off-by: Niklaus Giger <niklaus.giger@netstal.com>
16 years agoppc4xx: Beautify configuration files for Sequoia and Korat boards
Larry Johnson [Sat, 19 Jan 2008 02:49:05 +0000 (21:49 -0500)]
ppc4xx: Beautify configuration files for Sequoia and Korat boards

Signed-off-by: Larry Johnson <lrj@acm.org>
16 years agoMerge branch 'master' of git://www.denx.de/git/u-boot-arm
Wolfgang Denk [Fri, 15 Feb 2008 23:42:44 +0000 (00:42 +0100)]
Merge branch 'master' of git://denx.de/git/u-boot-arm

16 years agoWipe out assembler warnings while compiling x86 biosemu
Anatolij Gustschin [Fri, 15 Feb 2008 19:09:01 +0000 (20:09 +0100)]
Wipe out assembler warnings while compiling x86 biosemu

This patch tries to get rid of some assembler warnings about
changed .got2 section type while compiling x86 bios emulator
code.

Signed-off-by: Anatolij Gustschin <agust@denx.de>
16 years agoMerge branch '080208_dupint' of git://linux-arm.org/u-boot-armdev
Peter Pearse [Fri, 15 Feb 2008 13:00:54 +0000 (13:00 +0000)]
Merge branch '080208_dupint' of git://linux-arm.org/u-boot-armdev